Nonmonetary Transaction Fees

This is an attempt to address the objections to proof-of-stake made here: http://bitcoinmagazine.com/10829/impossible-trinity-security-environment-protection-decentralization/. This attempt applies the following two restrictions to the proof-of-stake model implemented by Peercoin:

[ol][li]Money created by block chaining is only useful in paying transaction fees.[/li]
[li]Only money created by block chaining is useful in paying transaction fees.[/li][/ol]

Here is a possible implementation:

[ol][li]If the private key controlling a stake output signs a transaction, then:
[list type=lower-alpha]
[li]This transaction must have its whole fee subtracted from that stake output.[/li]
[li]This transaction must have all its inputs coming from other, normal outputs.[/li]
[/list][/li]
[li]A private key controlling a stake output must sign every transaction.[/li][/ol]

Those restrictions have the following advantages:

[ol][li]The rich no longer can become richer merely by chaining blocks: the money thus created is only useful in paying for additional transactions, which increases fairness.[/li]
[li]A network-related activity (chaining blocks) has a network-related (making transactions) rather than monetary reward, which increases consistency.[/li]
[li]Chaining blocks becomes a requirement for making transactions, instead of needing some incentive, which increases network security by enforcing decentralization.[/li]
[li]The ratio between the block-chaining reward (stake output) and individual transaction fee only means the maximum number of transactions allowed by the network, which reduces its adjustment to a problem of network administration rather than monetary policy.
[/li][li]No block-chaining reward (stake output) or transaction fee has any monetary value left, which keeps the money supply unaffected.[/li][/ol]

[quote=“mirelo, post:1, topic:1765”]This is an attempt to address the objections to proof-of-stake made here: http://bitcoinmagazine.com/10829/impossible-trinity-security-environment-protection-decentralization/. This attempt applies the following two restrictions to the proof-of-stake model implemented by Peercoin:

[ol][li]Money created by block chaining is only useful in paying transaction fees.[/li]
[li]Only money created by block chaining is useful in paying transaction fees.[/li][/ol][/quote]

Do you propose an account/ledger/color-in-the-blockchain to keep track of money created by block-chaining, since now it is a separate money and can only be used to pay tx fees?
Does everyone have to create his/her own tx fee by doing block-chaining work? If not, how to let A use the money created by his block-chaining to pay B’s transaction fees?

With nonmonetary transaction fees, a stake output has no monetary value, so it makes no sense to use regular coins to represent it. Instead, I see two options, depending on whether we take transaction size into account or not:

[ol][li]A stake output directly represents the maximum number of transactions it allows its owner to make: each transaction signed by the private key controlling that stake merely subtracts one unit from it.[/li]
[li]A stake output indirectly represents that maximum number of transactions, by representing their total size: each transaction signed by the private key controlling that stake subtracts its own size from it.[/li][/ol]

Unfortunately, at least for now, I do not have enough knowledge of the protocol, transaction and block formats, and source code to suggest an implementation.

To sign a transaction, the private key controlling a stake output and that controlling each input of this transaction need not have the same owner. Let us imagine Bob has just received his first coins and has no time to wait until chaining his first block to make a transaction. On the other hand, Alice has more stake output than she needs. Bob can sign his transaction with the private key controlling each of its inputs and send this transaction to Alice, who then can sign the same transaction with the private key controlling her excessive stake output.

However, for this to become widespread, people willing other people to sign their transactions must pay them a monetary fee (in regular coins). The mechanism for this can be similar to the mechanism by which Bitcoin transactions pay fees to miners. We can probably use Bitcoin smart fees algorithm, when it is ready.

The whole point is: those monetary fees remain complimentary to the fundamental mechanism of securing the network, which is nonmonetary fees. People will always have the option of doing their own block chaining to make their transactions for free. So the resulting market for signing other people’s transactions would never be a security threat: in the long run, it tends to disappear, with the network approaching total decentralization.